Abstract The centralization of data management and statistical analyses for multimodal imaging studies facilitates a consistent approach to the identification of biomarkers and mechanisms for disease. The first goal of the Biosta- tistics and Neuroinformatics (BNI) core is to provide a resource for investigators to develop consistent, robust statistical designs for testing the results against hypotheses, including developing statistical models for the output of new analysis algorithms or summary measures, and protecting against replication issues. The second goal is to provide a) a foundation for data management and integration across modalities, making it straightfor- ward to identify and retrieve specific scans, b) quality assurance measures, and c) clinical assessments for the same subject or subsets of subjects, while interacting flexibly with the advanced algorithms and methods from the other cores. This same infrastructure is used as an ongoing resource to share the data as well, publicizing the availability of the data and allowing the research community to request and download the data for re-use. These two goals interact in providing ongoing support in the development of new pilot projects as the COBRE progresses, and in providing training opportunities in biostatistical issues specific to these complex neuroimag- ing, longitudinal, data-fusion analyses. The long term goal of this core is to provide a powerful resource for data capture, sharing and dissemination, and in collaboration with the other COBRE cores, to support the develop- ment of a diversified neuroimaging research environment that will continue to be competitive nationally and in- ternationally. 2